Measuring the luminosity at LHCb

Most physics analyses require that the luminosity, L, of the beam is know (i.e. the number of particles per cm2 per second). The higher the luminosity, the greater the number of events, N, in any physics process. The probability of any physics process happening (called the cross-section, s), is found by counting the number of events seen and dividing by the luminosity: s=N/L.

Traditional techniques to measure the luminosity using the beam current and size give a precision of about 10%. Some effort at LHCb has gone into a technique that uses beam gas collisions . Another method is to measure the number of events of a process with a precisely known cross-section. However, in the new energy regime of the LHC, very few processes are theoretically known to high accuracy. One candidate is the Z boson production rate (theoretically known to 3%). However since we want to measure this , we need to find another process. A recent paper by Shamov and Telnov (re)proposes the use of two-photon production of pairs of dimuons. The original theoretical paper suggests a precision better than 1% and although it suffers from a low cross-section it can calibrate any stable high-rate monitor.

We have undertaken a study to see if LHCb can trigger on these events, to estimate the statistical uncertainty and identify the limiting experimental systematics for this channel. We have interfaced the LPAIR generator to the LHCb code and find that with suitable cuts the signal can be extracted quite cleanly, with a background of about 10%. Most of that background comes from pion mis-identification, which can be estimated from the data itself. First results have been shown at the LHCb Production and Decay Working Group.

Here is Jonny's plenary talk in the LHCb week, June 6th.

Following on from many discussions with theorists on the various other sources of backgrounds, we released a LHCb note that contains a lot of details on how to measure the luminosity.